Purpose I was sitting in a coffee shop near conference hall about half hour after my team's submission of paper on GAIT ANALYSIS and explaining audience our method of approach. The feeling was ecstatic, but I was also a bit nervous as to know the result of the conference. Yet, it was satisfying to reach our goal after much of the struggle during initial days trying to decode lot of research papers to study methods which were used previously. p reviously. Sipping Sipping coffee, I gathered my mind to look back and check why I made decision to take this important career step . The day when I first visited a super specialty hospital, I had seen various devices capturing various signals of our body like EEG, ECG etc.. I had wondered what those plots gathered just by placing electrodes could provide information to a Doctor to infer status of a patient. The above incident kept lingering in my mind only to be back a few years later when I visited NIMHANS, Bangalore to fetch EEG signals from the patients with my sister for her undergraduate undergraduate project , she was assisting assisting the doctor in placing them in specific positions and he was explaining waveform that he obtained and how he it can help diagnose diseases from it. I was standing there spell bounded realizing that these are produced as result of the activity of millions of neurons in the brain and a waveform from them reflecting the patient's status was something that amazed me, appreciating the importance of those waveforms played in his day to day life, and at that instant I decided to consider seriously about this as a career. Not until fourth semester in Engineering Engineering I knew the “waveform” that I had speaking speaking of was a signal and differed differed a great deal from waveform. waveform. Signals and Systems Systems course It was, which made me understand this. It also learnt great deal about the role Mathematics plays in study of these . Digital Signal Processing followed it providing a practical framework to operate work upon on these signals on computer. But it is Real time signals which are random which proves difficult and all the above methods discussed considered signals to be Linear and Time Invariance which is completely violated by real time signals so a new frame work under probability theory has to be developed . I have particular reminiscence reminiscence of post classroom sessions in Linear algebra algebra which have always proved a valuable assets during my learning stages of these subjects. I have an inclination towards both theoretical and application oriented learning. My research interests lie in the fields like Speech processing, Image processing, Estimation and Detection. During my college term I had an privilege to learn Speech, Image and Biomedical signal processing as a part of my curriculum which was offered as an elective. Here I studied about the fundamentals on these subjects also worked on various assignments as a part understanding the subject in subtler way. This practical approach helped me in grasping the subject thoroughly as days progressed during my final year project. Particularly I was more interested in probabilistic methods of these Speaker recognition using methods like HMM and I found ways to learn these on my own. In fact, learning the methods and implementation of this by doing a project and aim to form a MATLAB Toolbox on the same.
Furthermore, the interaction with professors and participants during the conferences and the project exhibit conducted by various universities and colleges have always captured my interest and provided motivation to work on projects during my 2nd year of Engineering. Particularly, Our college had conducted 3 workshops, spanning a week , under the banner of Center of Pattern Recognition. Workshops, along with covering modules in conference such as Fundamentals of Signal Processing, Optimization techniques, Fundamentals of Pattern Recognition and Image processing , had provided Mathematical basis for the core concepts. Automated Obstacle Detector in Train an 8051 based project was a mini project which we . This aimed to solve problem manual signal crossing using the feedback from the infrared sensor. With further extension intended to include sophisticated modules on our train such as fire safety by placing a smoke sensor the project was completed in leisure. I had simulated a BPSK and QPSK over AWGN channel in MATLAB/Simulink, though it was not something new, the attempt was to learn more about Simulink through a project. During this phase I realized the importance of self learning, it made me correct my conceptual understanding of the subject. Aim of this exercise was to particularly get accustomed with programming in MATLAB/Simulink. The sense of satisfaction derived from this gave me more confidence to take on my future endeavors. During the final year of under graduate studies, as a part of project and seminar, I took the my seminar topic as Support Vector Machines partly. From my previous encounters doing the project, the strong belief was that, one has to learn Mathematical aspects behind the topic along with the small presentation to illustrate the application which invokes interest in audience. The preparation for the seminar was particularly tedious one since I lack basics to understand the subject thoroughly. Approaching the Dr Gouri Srinivasa PhD, in Information science department proved particularly enriching getting the overall picture of the content. Even then, I strongly felt the need to address the concepts of Hyperplanes in linear algebra (I understood the importance of abstract properties of planes when extended to more generalized case),Quadratic Optimization and Kernels used in Nonlinear classification myself. Though I could not understand these concepts to greater depth the outcomes from these are very prominent which only enhanced my interest in the learning more theoretical aspects of the subject. The seminar was appreciated by one and all in the department. In a way this path led me to know the nuances of the field in greater detail. The experience was similar in doing our project under Prof. Subhash .K .Kulkarni, being the lead of the group, The idea was to attack the problem of analyzing Gait of a person using the level sets. Silhouette of a particular person was extracted with background subtraction performed on the image to obtain the object under consideration. Next, we transformed the image obtained to get measures required for our calculation to obtain the signatures that are scale and rotation invariant. The result was a success for still images and needed a further sophisticated approach to extend the method for the video and real time application. Having contributed in discovering and evolving 2 of the 3 signatures we obtained, I contributed towards implementation of the obtaining level set contour image on MATLAB. My team produced a paper in International Conference in Signal and
Image Processing conference titled Novel Signatures for Gait Analysis Using Level Sets. (http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6754883)
During
one
of
the
conversation with the Prof. Ramasubramaniam, Ph.D,a prominent researcher in speech processing from IISc,Bangalore, I realized the importance for statistical classifying methods in the Pattern Recognition methods on images during the real time applications. This spurt my desire towards learning the statistical methods used in signal processing applications . During the tenure of the project, I had an opportunity to work as a Research Assistant under the my guide for about 10 months. Where I worked in developing paper also taught in labs like analog electronics, Digital Signal Processing for about two months . The experience was enriching which has captured my mind and made my mind to pursue my career in Research . My Career Goal: •
I want to see myself as an expert with deep knowledge with solid mathematical
background and taking small steps in developing my own algorithms with complete knowledge to understand and use most signal processing and communication algorithms. •
In the immediate future, I wish to explore Industry to train myself in applications of
Practical importance. •
I would like to take up Doctoral studies in about 4years with a major in mathematics and
concentrate on open ended problems in communication and cognitive science. I am following University of Colorado-Boulder(UCB) since MAVEN exploration satellite came in to foray in which UCB played a prominent role. The university has a strong association with NASA and has produced 18 Astronauts most of them are the alumni of this University. Along with JPL, UC-Berkeley, Caltech, LASP in UCB had a strong presence in research activities of NASA. Always producing best of the researchers to the scientific community. Number of Nobel Laureates and huge number of members of Prestigious Academic academies are the indication of strong research culture that exists in the University. I would like to contribute to the university's Flagship program Vision 2030 which aims to be one among the top research communities in the world with excellent infrastructure. T Bridges the program which aims to get older adults to mainstream technologies and make them aware of is an innovative way to bring technology and community service in tandem to work for betterment of the society which I always wanted to do. Having realized the importance of Entrepreneurship, employment and knowledge generation the university has taken several steps to scale it to greater heights by developing funding resources and innovation centers along with Management Department. STARTUP COLORAD, SPARK BOULDER, Engineering entrepreneurship Certificate program from LEEDS SCHOOL OF BUSINESS are only few of many initiatives taken in this regard. I want to explore the possibility of establishing a firm using the knowledge and expertise I possess in my field. SWARM ROBOTICS-Intelligent robots which
work in tandem in performing complex tasks such as Bees and algorithms developed in this regard impressed allot on the level of research going on in the campus.
The department of electrical(change) has some of the best labs and centers headed by prominent faculty in the domain. Specially in signal processing and communication which is my area of interest there are research groups headed by Dr. Tim Brown, Dr. Shannon Hughes, Dr. Youjian Liu, Dr. Mahesh Varanasi, Dr. Ute Herzfeld I yearn to learn from these professors and see myself nurturing my academic goals along with research ambitions. I want to be part of DSP and communications group which spearheads Pervasive Communication Lab under Dr Tim Brown since it is synchronous with my research interests. I have admired the work of Dr Hughes in her Detection of forgery in Paintings using supervised learning when I took machine learning course online as part of reading and also during Final year project in the college. She has authored a book in Modern signal processing which I had referred for the DSP course in the college. Her research areas like kernel methods in signal processing and manifold learning methods in data analysis are the ones which I will be able to learn from and make contributions in the field. I have referred notes of Dr Youjian Liu in the courses like Linear Systems theory , Introduction to Probability in the college curriculum and a part of self study, which was informative and gave a different perspective to my learning. Prof. Liu who is a senior IEEE member who has research interests which partly aligns with mine like MIMO Communications, Space Time Coding, it will be an honor to earn a spot in his team. Trans disciplinary Spoken Language Research Lab which deals with speech recognition and synthesis has also contributed allot to the field it will be exciting to learn from their methods which enhances my research experience. I have tried to understand papers of these above mentioned professors and was just able to get superficial knowledge. I wish to develop algorithms and methods of my own and self learning process which I feel are utmost important for Doctoral Studies so courses offered such as Probability Theory, Stochastic processes,Real analysis, Digital Filters and Wavelets are the ones I would like to take up gain subtler understanding of the concepts. I am applying to UC Boulder because for its emphasis on student research. The distinguished alumni are all around the world and are key resources in their respective domain. I am sure that with the learning and research gained from Masters Program will help me in future endeavors as a Doctoral Student. At this juncture, I hope and aim to make the best use of this exciting career path, through your reputed institute. With a staunch inclination towards problem solving, positive attitude and analytical assessment, I assure you that I will actively participate in the course you offer for Signal Processing and Communication. Possible inclusions :
Community services by me. Possible future technologies in signal processing.